Pulmonary Embolism and Deep Vein Thrombosis: Christopher Kabrhel

74
Pulmonary Embolism and

Deep Vein Thrombosis
Christopher Kabrhel
KEY CONCEPTS
• H allmarks of deep venous thrombosis (DVT) include unilateral limb pain and • P atients with low gestalt pretest probability and a negative PERC rule may
swelling, though these findings can be subtle and nonspecific. incur more harm than benefit if testing for PE (including D-dimer) is performed.
• Patients with low pretest probability (PTP) can have DVT ruled out in the ED • Patients with non-high pretest probability (e.g., Wells Score ≤ 6) can have PE
with a negative D-dimer or venous ultrasound (US), patients with high PTP can ruled out with a negative D-dimer.
have DVT ruled out in the ED with a negative D-dimer and venous US. • The D-dimer threshold can be adjusted according to the patient’s age or the
• Three-point US evaluates the leg veins proximal to the knee and has a sen- pretest probability of PE. The formula for age adjustment is Age × 10 ng/mL.
sitivity of 95% and specificity of 95% for proximal DVT when performed by Using an adjusted D-dimer threshold reduces the need for imaging by about
a certified sonographer. A negative three-point ultrasound in a patient with 10% to 15%.
a moderate or high PTP for DVT should have a D-dimer test or repeat venous • Patients with high pretest probability or a positive D-dimer require imaging.
ultrasound within 7 days. A single, whole-leg ultrasound excludes DVT in all For most patients, including pregnant women, computer tomography pulmo-
pretest probabilities. nary angiography (CTPA) is the imaging test of choice.
• Anticoagulation for DVT and pulmonary embolism (PE) can be achieved with • For pregnant women, the decision to undergo any testing for PE should be
a direct-acting oral anticoagulant (DOAC), low-molecular-weight heparin shared with the patient. To minimize fetal radiation exposure, diagnostic test-
(LMWH), unfractionated heparin or, in rare cases, an alternative anticoag- ing for PE should start with a combination of D-dimer and bilateral lower
ulant. For DVT and most PE, DOACs are as effective as and safer than the extremity venous ultrasound.
combination of heparin and warfarin. • For patients diagnosed with PE, typical resuscitative measures can be harm-
• The treatment of isolated calf vein DVT is controversial, and it is reasonable ful. Endotracheal intubation and positive-pressure ventilation can decrease
to withhold anticoagulation in favor of a repeat venous ultrasound within 7 preload and precipitate cardiac arrest. Excessive intravenous fluid adminis-
days and close follow-up. tration can lead to worsening right ventricular distention and left ventricular
• Distal superficial vein thrombophlebitis can adequately be treated with non- compression.
steroidal antiinflammatory drugs and warm compresses, but proximal superfi- • Patients with PE should be risk stratified using a combination of vital signs,
cial vein thrombophlebitis should be treated with anticoagulation. CTPA, echocardiography, and troponin. High-risk PE is defined by hemody-
• Most patients with DVT distal to the iliofemoral region can be safely treated namic instability, intermediate-risk by right ventricular dysfunction, and low-
as outpatients as long as close follow-up and access to anticoagulant medi- risk by the absence of either.
cations is assured. • Patients with high-risk PE should be treated with thrombolysis or thromboem-
• PE can present asymptomatically, with dramatic clinical symptoms, or with bolectomy unless contraindicated.
sudden cardiac death. Even the most common symptom, dyspnea, is absent in • Patients with intermediate-risk PE are usually not candidates for thrombolysis
a quarter of patients. but may be candidates for catheter-directed or low-dose systemic thromboly-
• The first step in the evaluation of possible PE is determining whether testing sis.
for PE is indicated. • About 25% of patients with low-risk PE can be treated as outpatients.
• The patient’s pretest probability for PE, as determined by clinical gestalt or
a validated score (e.g., the Wells Score) dictates the approach to objective
testing for PE.
FOUNDATIONS older patients have a much higher incidence, at 1/10,000 in people 20

to 30 years old and 1/100 in people greater than 80 years old. How-
Background and Importance ever, even this high incidence grossly understates the importance of PE
Each year more than 500,000 people are diagnosed with venous throm- and DVT to the practice of emergency medicine. Because the clinical
boembolism (VTE), which includes deep vein thrombosis (DVT) and presentations of PE and DVT are highly variable, with symptoms and
pulmonary embolism (PE). PE is the third most common cause of car- signs that overlap with many common ED diagnoses, PE and DVT are
diovascular death in the United States. PE and DVT occur equally in frequently on the differential diagnosis for patients who present with
women and men, and while PE and DVT can occur in any age group, chest pain, dyspnea, syncope, tachycardia, hypoxemia, leg pain, edema,
1022
CHAPTER 74 Pulmonary Embolism and Deep Vein Thrombosis 1023
Coagulation ≤ Fibrinolysis Coagulation > Fibrinolysis who develop PE or DVT after exposure to heparin may have heparin-
induced thrombocytopenia, a hypercoagulable state leading to VTE.
While the presence of an occult risk factor may only become known
after a patient leaves the ED, a thorough history may suggest occult
cancer (e.g., unintentional weight loss) or recent exposure to heparin
Older age that would enable the emergency clinician to adjust their assessment of
Younger age Immobility a patient’s risk of PE or DVT.
Mobility Obesity
Normal BMI Cancer Anatomy, Pathology, and Pathophysiology of VTE
No comorbidity Inflammation
Low genetic risk Genetic risk The venous anatomy of the lower extremity is divided into the deep
Venous injury and superficial systems (Fig. 74.2). The superficial venous system con-
Prior VTE sists primarily of the greater and short saphenous veins and perforating
veins. Distal greater saphenous vein thrombi are often referred to as
superficial thrombosis, but greater saphenous clots near the connection
with the femoral vein should be referred to as DVT. The deep venous
system includes the anterior tibial, posterior tibial, and peroneal veins,
collectively called the calf veins. Venous thrombi isolated to the calf
veins are referred to as distal DVT. The calf veins join together at the
Fig. 74.1 Diagram of clotting in a lower extremity vein. Note how the knee to form the popliteal vein, which extends proximally and becomes
clot forms in and around the cusps of the valve. (Illustration courtesy of the femoral vein at the adductor canal. Venous thrombi in the popliteal
Sadie A. Kabrhel.) or more proximal veins are referred to as proximal DVT. The femoral
vein (previously known as the superficial femoral vein), is joined by
and other nonspecific complaints. The wide demographic range and the deep femoral vein and then the greater saphenous vein to form the
nonspecific clinical presentation of PE and DVT is why as many as 1 common femoral vein, which subsequently becomes the external iliac
in 10 emergency department (ED) patients are evaluated for possible vein at the inguinal ligament. Venous thrombi in the proximal femoral
DVT or PE. Emergency clinicians must, therefore, be knowledgeable and iliac veins are known as iliofemoral DVT.
and judicious in their approach to PE and DVT. Knowledge of the venous anatomy is also important for the perfor-
As illustrated in Figure 74.1, venous thrombosis occurs when the mance and interpretation of ultrasound of the leg veins. Compression
propensity of blood to coagulate overwhelms endogenous anticoag- ultrasound, including point-of-care ultrasound, is typically limited to
ulant and fibrinolytic systems. Numerous factors associated with the the common femoral, femoral, and sometimes popliteal veins. Duplex
classic triad of venous injury, venous stasis, and hypercoagulability ultrasound includes compression ultrasound of the proximal veins as
have been associated with an increased risk of VTE in epidemiologic well as Doppler ultrasound of the calf veins. The anatomic location of
studies (Table 74.1). Emergency clinicians should be aware of factors a DVT also influences prognosis because proximal DVT are less likely
that might increase a patient’s propensity to clot and should consider to lyse spontaneously, more likely to embolize, and are associated with
these factors when determining whether a patient’s clinical presenta- a greater risk of long-term complications such as post-thrombotic
tion warrants an evaluation for VTE. Important factors include older syndrome.
age, prior history of VTE, active cancer, recent surgery or trauma, Although 90% of DVT form in leg veins, clots can also form in the
recent hospitalization longer than three days, limb immobility, and arm, around venous catheters, pacemaker wires, or other foreign bod-
estrogen use (especially if initiated in the past three months). Thus, the ies such as inferior vena cava filters. Other sites of venous thrombosis
age and comorbid illness of the population served by an ED determine rarely encountered in the ED include the jugular, ovarian, mesenteric,
the frequency of VTE diagnoses. renal, portal, hepatic, cerebral, and retinal veins.
However, not all epidemiologic risk factors are associated with DVT formation typically begins when monocytes expose blood to
an increased likelihood of an ED diagnosis of VTE (Table 74.1). For tissue factor on their surfaces. This process overwhelms natural anti-
example, pregnant women are five times as likely to develop VTE as coagulant and fibrinolytic mechanisms and leads to the aggregation of
nonpregnant women of the same age, but only 4% of pregnant women red blood cells, platelets, and fibrin in the venous sinuses or cusps of
tested for VTE in the ED have the diagnosis, compared to 12% of non- the lower extremity deep veins. The formation of a DVT causes vas-
pregnant patients. This is because emergency clinicians have a low cular congestion which, in turn, causes veins to dilate and valves to
threshold to test pregnant patients for VTE. Similarly, travel is associ- become incompetent. The lack of forward venous blood flow leads to
ated with a small increase in the absolute risk of VTE, but a history of venous stasis and further thrombus propagation.
travel often leads emergency clinicians to test for VTE, so travel does When a DVT embolizes, it flows proximally through the venous
not seem to be associated with VTE in patients tested in the ED.1 The system toward the vena cava and the heart. In 3% of cases, a por-
decision to initiate an evaluation or treatment for VTE should be based tion of the clot will remain in the right atrium or ventricle, a condi-
on an analysis of the risk PE poses to the patient weighed against the tion known as clot-in-transit. Clots that pass through the heart and
risk imparted by testing and treatment. enter the pulmonary arterial circulation are called PE. A PE can be
As many as 50% of patients diagnosed with PE have no apparent described as saddle if the clot is visualized across the bifurcation of
clinical risk factors for PE or DVT at the time of diagnosis. However, the main right and left pulmonary arteries (Fig. 74.3A). More distal
the lack of a risk factor known at the time of the ED diagnosis does not PE are typically described according to their anatomic location. Clots
mean that an occult risk factor does not exist. Between 4% and 11% of may lodge in a main pulmonary artery (Fig. 74.3B), or in a lobar (Fig.
patients with PE or DVT will receive a new cancer diagnosis within 1 74.3C), segmental (Fig. 74.3D) or subsegmental pulmonary artery
year of VTE diagnosis, depending on whether the VTE event is pro- branch. PE may extend from a proximal artery into distal branches,
voked or idiopathic. Genetic or acquired hypercoagulable states are and PE frequently fragment, lodging in multiple arterial branches
often unknown until a patient has their first thrombotic event. Patients simultaneously.
1024 PART III Emergency Medicine by System
TABLE 74.1 Epidemiologic Risk Factors and Physiologic Findings for Pulmonary Embolism in
the Emergency Department Setting
Strength of Association With PE or DVT
Risk Factor from Epidemiologic Study Mechanism in ED Patients
Surgery Inflammation ++++
(within past 4 weeks, requiring general anesthesia) Venous Injury Stasis
Trauma within past 4 weeks requiring Inflammation +++
hospitalization Venous Injury Stasis
Older age Hypercoagulability +++
Stasis
Association with other diseases of aging (e.g., cancer,
surgery)
Prior PE or DVT Hypercoagulability ++
Stasis (due to valvular damage)
Active cancer Hypercoagulability ++
Inherited or acquired thrombophilia Hypercoagulability ++
Limb or generalized immobility Stasis ++
Estrogen Hypercoagulability ++
Pregnancy or postpartum Hypercoagulability +
Recent travel Stasis Minimal
Connective tissue disease Inflammation Unknown
Inactive cancer (considered in remission) Hypercoagulability Not significant
Smoking Inflammation and hypofibrinolysis Not significant
Family history of VTE Hypercoagulability (inherited predisposition) Not significant
Symptoms Mechanism Strength of association with PE or DVT in ED Patients
Hemoptysis Infarction +++
Pleuritic chest pain Lung infarction +
Dyspnea V/Q mismatch +
Syncope Obstruction of pulmonary outflow tract +
Sudden onset of symptoms Vascular obstruction Not significant
Substernal chest pain Presumed cardiac ischemia Not significant
Signs
Unilateral leg or arm swelling Venous obstruction ++++
Unexplained hypoxemia (Sao2 < 95% [sea level]) V̇ / Q̇ mismatch +++
Pulse rate > 100 beats/min Cardiac stress, baroreceptors ++
Note: Differences between epidemiologic risk factors and risk factors for a diagnosis of PE or DVT in the ED reflect the ability of these factors to
differentiate patients who present with symptoms consistent with PE or DVT who are ultimately diagnosed with PE or DVT from those who have
PE or DVT excluded after their ED workup.
PE, Pulmonary embolism; DVT, deep vein thrombosis; ED, emergency department; V̇ / Q̇ , ventilation/perfusion ratio; Sao2, saturation of oxygen mea-
sured on pulse oximetry.
The physiologic effect of an obstruction in the pulmonary vascula- the systolic RV is crescentic in cross-section and the left ventricle
ture is variable. Small, subsegmental PE often lyse spontaneously and (LV) is circular. However, with increased PA pressures, the systolic RV
may be clinically inconsequential. Subsegmental thrombi that do not dilates asymmetrically towards the intraventricular septum, compress-
spontaneously lyse but obstruct blood flow to the lung’s periphery can ing the LV into a “D” shape on cross-section. This can lead to underfill-
cause pulmonary infarction, necrosis, pleural inflammation, and severe ing of the left ventricle, decreased cardiac output, decreased coronary
pleuritic pain. The right ventricle normally pumps through a pulmo- artery perfusion, myocardial ischemia, circulatory collapse, and death.
nary vascular tree with a low resistance to flow, and young individuals
without cardiopulmonary disease can tolerate at least 30% obstruction DEEP VEIN THROMBOSIS
from a clot with minimal symptoms or signs. Larger PE can cause an
acute increase in pulmonary artery pressure due to a combination of Clinical Features
mechanical obstruction and chemically mediated vasoconstriction of Hallmarks of DVT include unilateral limb pain and swelling, though
unobstructed pulmonary arteries. Increased right ventricular (RV) these findings can be subtle and nonspecific. Patients may report only
afterload (i.e., when the pulmonary artery systolic pressure exceeds 40 mild cramping or a sense of fullness in the calf. The clinical signs of
mm Hg) can lead the thin-walled right ventricle to dilate and become DVT may include edema, erythema, and warmth of the affected
hypokinetic. As illustrated in Figure 74.4, with normal PA pressures, extremity, tenderness to palpation along the distribution of the deep
Inferior vena cava muscles causes inflammation that appears clinically similar to DVT.
Common iliac vein Spontaneous calf muscle hematomas can also cause pain and inflam-
matory changes. Many patients with systemic edema (e.g., from heart
Internal iliac vein
External iliac vein failure) will have asymmetrical swelling that raises concern for DVT.
Common femoral vein
Diagnostic Testing
Greater saphenous vein Pretest Probability Estimation
Diagnosis of DVT (and PE) starts with an estimation of the pretest
Deep femoral vein
probability (PTP). PTP estimation helps guide the choice of diagnostic
tests, the interpretation of results, and the need for follow-up testing.
Femoral vein However, in the eyes of many clinicians the noninvasive, nonionizing
nature of testing for DVT makes PTP estimation for DVT less critical
than it is for PE. PTP estimation may be accomplished by the clini-
cal gestalt of an experienced clinician or in conjunction with a clinical
Popliteal vein decision tool.
The most commonly used and well-validated clinical decision tool
Gastrocnemic vein
for DVT is the Wells DVT score (Table 74.2). Because the components
of the Wells DVT score have been incorporated into clinical gestalt over
Soleus vein the years since the score was developed, Wells DVT score and clinician
gestalt have approximately equal diagnostic accuracy, and either method
Posterior tibial vein is acceptable. Although the Wells DVT score can categorize patients as
low, intermediate or high PTP, the decision to test with a highly sensi-
Anterior tibial vein
Peroneal vein tive D-dimer or venous US is dichotomous. Therefore, it is easiest and
appropriate to combine low and intermediate probability groups and
consider the PTP of DVT to be low (−2 to 2 points) or high (≥3 points).
The Wells DVT score has not been validated in pregnant women,
but the LEFt score has been validated in a study of 157 pregnant
Fig. 74.2 Diagram of the leg vein anatomy relevant to the diagnosis of women. It consists of 1 point in case of left (L) leg suspicion, 1 point
deep vein thrombosis. A three-point ultrasound includes the common
for edema (E), and 1 point if the suspicion occurred during the first
femoral, femoral and popliteal veins. A whole-leg ultrasound adds the
greater saphenous, posterior tibial, peroneal veins and the gastrocne-
trimester (Ft) of pregnancy. A LEFt score of 0 or 1 indicates low PTP.
mius vein. (Illustration courtesy of Sadie A. Kabrhel.) Although not validated, an approach that substitutes the LEFt score for
the Wells score in pregnant women is reasonable.
venous system, dilation of superficial collateral veins, and rarely a pal- Laboratory Evaluation
pable venous cord. Figure 74.5A shows a patient with a left leg DVT.
The D-dimer test measures the enzymatic breakdown of cross-linked fibrin
To illustrate the difficulty of differentiating DVT from alternative
from any intravascular thrombus. A normal quantitative D-dimer con-
diagnoses based on clinical examination, Figure 74.5B shows a patient
centration in a patient with a low PTP (e.g., Wells score −2 to 0) excludes
with a ruptured Baker cyst in the left leg. Fever suggests an alternative
proximal DVT with a sensitivity of approximately 92% and a specificity of
diagnosis, such as cellulitis. Because the left iliac vein is vulnerable to
45%.2 This sensitivity is slightly lower than the sensitivity for PE, possibly
compression by the left iliac artery (May-Thurner syndrome), leg DVT
because DVT tend to be subacute, and therefore less prone to release D-di-
occurs with a slightly higher frequency on the left. Bilateral leg DVT is
mer, when diagnosed. Box 74.2 lists conditions other than PE and DVT
found in fewer than 10% of ED patients diagnosed with DVT.
that elevate the D-dimer. The specificity, and therefore the clinical useful-
More than 90% of upper extremity DVT occur in the presence of an
ness, of the D-dimer may be low in patients with these conditions.
indwelling catheter or similar device. Therefore, arm pain or swelling in
The US Food and Drug Administration (FDA) has cleared numer-
the same arm as a catheter, infusion device, or pacemaker wire should
ous D-dimer assays to aid in the diagnosis and exclusion of VTE,
raise suspicion for DVT. In the absence of a device, upper extremity
mostly using a cutoff of greater than 500 ng/mL to define abnormal.
DVT tends to occur in the dominant arm of young athletes, a condition
However, some D-dimer assays have cutoffs other than 500 ng/mL, so
known as Paget-Schroetter syndrome. Paget-Schroetter syndrome is an
it is important for emergency clinicians to know the threshold consid-
effort-induced form of thoracic outlet syndrome. Repetitive motion of
ered positive in their practice setting.
the arm in the setting of hypertrophied scalene muscles or congenital
Studies find that the need for venous US can be decreased by 5% by
cervical ribs causes compression of the subclavian vein and DVT.
using a threshold for a positive D-dimer that is adjusted according to
Differential Diagnosis the patient’s age using the following formula:2-5
Box 74.1 lists differential diagnoses for DVT. Venous insufficiency that Age × 10 ng / mL

causes congestion and inflammation is a common alternative diagno-
sis for DVT, especially since venous insufficiency increases the risk Thus, an 80-year-old patient with a PE unlikely or non-high PTP can
of DVT. Cellulitis is another common consideration. However, in a have DVT (or PE) excluded with a D-dimer concentration less than
patient with clinical evidence of cellulitis, the frequency of concurrent 800 ng/mL. This strategy maintains a diagnostic sensitivity near 95%
DVT is only about 3%. Injuries to the gastrocnemius muscle or Achilles but increases the percentage of patients who can have DVT excluded
tendon can mimic the pain and swelling of DVT but are usually dis- without venous ultrasonography. The safety of this strategy has not
tinguished by history. Enlargement of the synovial membrane (Baker been tested with D-dimer assays with abnormal thresholds different
cyst) can cause popliteal pain and rupture of the cyst into the calf than 500 ng/mL.
A B
Saddle PE Main Pulmonary Artery PE
C D
Lobar PE Segmental PE
Fig. 74.3 Diagnostic computed tomography pulmonary angiography (CTPA) scans showing: (A) saddle PE,
(B) right main pulmonary artery PE, (C) right lobar PE, (D) right segmental PE. The pulmonary arteries are
enhanced by contrast, but PE appear as filling defects, as indicated by the yellow arrows.
An alternative to age adjustment is to apply a Bayesian approach, radiologist, has a sensitivity of 95% and specificity of 95% for DVT. A
by which the threshold for a positive D-dimer is increased based on negative three-point venous duplex ultrasound excludes the diagnosis
PTP. Using this strategy, patients with low PTP can have DVT ruled out of DVT in patients with low PTP. However, for patients at high risk,
using a threshold of 1000 ng/mL (for a test that normally uses 500 ng/ a negative three-point ultrasound is inadequate to exclude DVT as a
mL). This approach has been shown to safely increase the proportion of single test. Patients with high PTP can have DVT ruled out in the ED
patients with a negative D-dimer results from 50.9 % to 56.1%.6 with a combination of a negative three-point ultrasound and a nega-
tive D-dimer. A patient with high PTP, an elevated (or not performed)
Radiographic Evaluation D-dimer, and a negative three-point ultrasound should be referred for
An expertly performed and interpreted positive venous ultrasound is suf- a repeat venous US in 5 to 7 days to ensure their symptoms are not due
ficient to confirm the diagnosis of DVT. Several approaches to performing to a distal (calf vein) DVT that later propagates proximally.
venous US are commonly used, and the emergency clinician should be Bedside point-of-care ultrasound (POCUS) performed by a trained
aware of the US protocols used in their practice setting. Whole-leg US is emergency clinician is 90% to 95% accurate compared to radiology-
the criterion standard and combines compression US of the proximal veins performed US.7-9 Test characteristics are highest for femoral veins and
with ultrasound of the calf and saphenous veins. Whole-leg US is associ- lowest for saphenous and popliteal veins. POCUS of the upper extremity
ated with a venous thromboembolism event rate at 3 months of 0.5%. A is not well studied and should be performed by radiology when necessary.
single normal whole-leg ultrasound excludes DVT regardless of PTP. Ultrasound cannot be used to rule out iliac or pelvic vein thrombo-
Three-point US, also called compression US, images the common sis. For this, venography (typically CT) is needed. When duplex ultra-
femoral, femoral, and popliteal veins. Three- point US, when per- sound is not available, the decision to empirically anticoagulate while
formed by a certified sonographer and interpreted by a board-certified awaiting the availability of ultrasound imaging should be based on the
PTP of DVT, and the risk anticoagulation poses to the patient. Gener- high risk for pelvic vein thrombosis (e.g., a patient with gynecologic
ally, patients with low PTP do not need empirical anticoagulation while malignancy and bilateral leg swelling). Its use is limited by cost, avail-
they wait for diagnostic imaging. ability, and patient tolerance. In only about 20% of patients with pelvic
Magnetic resonance imaging (MRI) can evaluate the pelvic vascu- DVT is the clot isolated to the pelvic veins. Therefore, even when pelvic
lature and inferior vena cava (IVC), which is not possible with ultra- vein DVT is suspected, venous US is still the recommended first test.
sound. Although CT venography can also visualize the pelvic veins Figure 74.6 provides a diagnostic flowchart for DVT that incorpo-
and IVC, the accuracy varies across studies, and CT exposes patients rates PTP. Patients with low PTP can have DVT ruled out in the ED
to ionizing radiation. MRI does not produce ionizing radiation. Thus, with a negative D-dimer or venous US. Patients with high PTP can
MRI is a reasonable option to evaluate the pelvic veins of patients at have DVT ruled out in the ED with a negative D-dimer and venous
US. Patients with high PTP, a positive D-dimer, and a negative venous
US of the proximal veins should have a follow-up ultrasound one week
after ED discharge to look for the propagation of a distal DVT that
might not have been detected during the initial ED visit.
Management
Patients with a positive ultrasound and patients with high PTP for whom
Normal PA pressure Normal PA pressure imaging will be delayed (e.g., until the next day) should have anticoagu-
Diastole Systole lation initiated in the ED at the time of diagnosis, unless contraindicated.
Therapeutic anticoagulation for PE and DVT is the same. Options
for initial anticoagulation are listed in Table 74.3. Direct-acting oral
anticoagulants (DOACs) are as effective as warfarin at preventing
recurrent VTE and are associated with fewer bleeding events, and
in particular, intracranial bleeding events. They are well tolerated by
patients and do not require injections or monitoring. The DOACs
Elevated PA pressure Elevated PA pressure BOX 74.1 Differential Diagnosis for DVT
Diastole Systole
Fig. 74.4 Diagram showing the relationship of the right ventricle (RV) Venous insufficiency causing congestion and inflammation
and left ventricle (LV) under conditions of normal pulmonary artery (PA) Cellulitis
pressure and elevated PA pressure (i.e. in the setting of an acute PE). Muscle or tendon injury
Note how, in the bottom right illustration, when PA pressure is elevated Baker cyst (including ruptured synovial membrane)
the RV dilates asymmetrically compressing the left ventricular septum. Hematoma
This can lead to decreased filling, decreased cardiac output, myocar- Arterial insufficiency and claudication
dial ischemia, hypotension and death. (Illustration courtesy of Sadie A. Asymmetrical edema (e.g., due to congestive heart failure or liver disease)
Kabrhel.)
A B
Deep Vein Thrombosis Ruptured Baker Cyst
Fig. 74.5 Clinical photograph showing patients with (A) deep vein thrombosis of the left leg and (B) ruptured
Baker cyst of the left leg. Note the similar clinical appearance with both patients presenting with leg pain and
swelling.
rivaroxaban and apixaban do not require bridging with low-molecular- For patients with active cancer, LMWH is associated with a lower
weight heparin (LMWH) and are the first-choice anticoagulants for risk of VTE recurrence than warfarin, but emerging evidence sug-
most patients with DVT.10-12 gests that DOACs are also safe and effective in this population.13,14
Concern about the ability to reverse anticoagulation should not dis-
TABLE 74.2 Wells Score for Deep Vein suade an emergency clinician from starting DOAC therapy. Rever-
Thrombosis sal agents are available for all DOACs (Box 74.3), and bleeding that
does occur on DOAC therapy tends to be less severe than bleeding
Clinical Feature Score on warfarin.15,16 However, DOACs are either contraindicated or not
Active cancer (treated within the previous 6 months or 1 studied for the treatment of DVT associated with pregnancy, severe
currently receiving palliative treatment) renal failure, liver failure, antiphospholipid antibody syndrome, and
Paralysis, paresis, or recent plaster immobilization of the 1 high-risk PE.
lower extremities
Recently bedridden (for ≥3 days or major surgery within 12 1 BOX 74.2 Factors Other Than VTE
weeks requiring general or regional anesthesia) Associated With Positive D-Dimer Results
Localized tenderness along the distribution of the deep 1
venous system • Female sex
Entire leg swollen 1 • Advanced age
• Black or African American race
Calf swelling at least 3 cm larger than on the asymptomatic 1
• Cocaine use
side (measured 10 cm below the tibial tuberosity)
• Immobility (general, limb, or neurologic)
Pitting edema confined to the symptomatic leg 1 • Hemoptysis
Collateral superficial veins (nonvaricose) 1 • Hemodialysis
Previously documented deep vein thrombosis 1 • Malignancy (active)
Alternative diagnosis at least as likely as deep vein −2 • Rheumatologic disease (rheumatoid arthritis, systemic lupus erythemato-
thrombosis sus)
• Sickle cell disease
A score < 2 indicates that the probability of deep vein thrombosis is low. • Pregnancy and postpartum state
Adapted from: Wells PS, Anderson D, Bormanis J, et al. Value of • Recent surgery (within 1 month)
assessment of pretest probability of deep-vein thrombosis in clinical
management. Lancet. 1994;350:1795-1798. Note: Warfarin use is associated with false-negative D-dimer results
Assess pretest probability of DVT
Intermediate or
Low
high
(Wells <2)
(Wells 2+)
Three-point Whole-leg Three-point

D-dimer +
venous ultrasound venous ultrasound venous ultrasound
–
– +
– –
D-dimer
–
Repeat venous
US in 7 days
No DVT DVT
Fig. 74.6 Flowchart algorithm showing the diagnostic approach to a patient with possible acute deep vein
thrombosis. DVT, Deep vein thrombosis; −, test negative; +, test positive; US, ultrasound.
TABLE 74.3 Anticoagulant Options for the Initial Treatment of DVT or PE

Anticoagulant Initial Dose Restriction Time to Peak
Unfractionated heparin 70–80 U/kg, then 17–18 U/kg/h, IV Heparin-induced thrombocytopenia 1 hour
Enoxaparin 1 mg/kg q12h or 1.5 mg/kg q24h Creatinine clearance < 30 mL/min 3 hours
subcutaneously
Dalteparin 200 unit/kg daily or 100 unit/kg q12h Creatinine clearance < 30 mL/min 4 hours
subcutaneously
Fondaparinux 5–10 mg subcutaneously Creatinine clearance < 30 ml/min 3 hours
Rivaroxaban 5 mg by mouth q12h for 21 days with Creatinine clearance < 30 mL/min 2–4 hours
food
Apixaban 10 mg by mouth q12h × 7 days Creatinine clearance < 30 mL/min 3–4 hours
followed by 5 mg by mouth q12h with
or without food
DVT, Deep vein thrombosis; PE, pulmonary embolism.
BOX 74.3 Reversal Agents for BOX 74.4 Contraindications to

Anticoagulants Anticoagulation
Anticoagulant Reversal Agent Absolute contraindications to anticoagulation include:
• Active bleeding into a critical organ or uncontrolled site
Heparin Protamine sulfate
• Severe bleeding diathesis
Warfarin Fresh frozen plasma
• Recent, planned, or emergency high-bleeding-risk surgery or procedure
4-Factor prothrombin complex
• Recent major trauma
concentrate
• Recent intracranial, spinal or ocular hemorrhage
Vitamin K
Relative contraindications to anticoagulation include:
Dabigatran Idarucizumab
• History of gastrointestinal major bleeding
Rivaroxaban, Apixaban Andexanet alfa
• Intracranial or spinal tumors
• Previous bleeding into a tumor
• Large abdominal aortic aneurysm with concurrent severe hypertension
Thrombolysis, whether systemic or catheter-directed, for DVT not
• Stable aortic dissection
associated with limb ischemia has not been shown to improve mor-
• Recent, planned, or emergent low-bleeding-risk surgery/procedure
tality or post-thrombotic syndrome (pain, paresthesia, induration,
swelling, discoloration, and ulceration of the leg after DVT) but does
increase the risk of bleeding.17,18
Compression stockings can no longer be advocated routinely for BOX 74.5 The VTE-BLEED Score
DVT, as the data do not support a reduction in post-thrombotic syn- • Active cancer, 2 points
drome. However, the reduction in post-thrombotic syndrome may • Male patient with uncontrolled hypertension, 1 point
depend on how soon after DVT diagnosis the patient starts wearing • Anemia, 1.5 points
the stockings, and some studies do show an improvement in quality of • History of bleeding, 1.5 points
life, so some patients may find them beneficial.19,20 Patients should be • Renal dysfunction (creatinine clearance 30–60 mL/min), 1.5 points
encouraged to ambulate after anticoagulation for DVT to reduce the • Age ≥60 years, 1.5 points
incidence of post-thrombotic syndrome.
Assessing Bleeding Risk with superficial thrombophlebitis involving the greater saphenous vein
Prior to initiation of anticoagulation for VTE, emergency clinicians that extends above the knee are at risk for progression to DVT via the
should assess the patient’s risk of bleeding. This is particularly important saphenous-femoral vein junction. In these patients, an abbreviated 45-
for patients with calf vein DVT without PE or isolated subsegmental PE day course of prophylactic-dose anticoagulation reduced clot exten-
without DVT, for whom withholding anticoagulation may be reasonable. sion, PE, and DVT, though the absolute rates of PE were less than 1%
Absolute and relative contraindications to anticoagulation are listed in Box regardless of treatment allocation. If a greater saphenous vein clot is
74.4. Bleeding risk can also be assessed using the validated VTE-BLEED proximal (within 3 cm of the connection with the femoral vein; see Fig.
score (Box 74.5). Patients with less than 2 points have a 0.5% risk of major 74.2), the risk of extension to the deep venous system is about 25% so
bleeding, compared to 2.0% in patients with scores of 2 or more points.21-23 therapeutic (full dose) anticoagulation is warranted for at least 30 days.
Superficial Vein Thrombophlebitis Isolated Calf Vein Thrombosis

Distal superficial vein thrombophlebitis can adequately be treated with The optimal management strategy for thromboses isolated to the calf
nonsteroidal antiinflammatory drugs and warm compresses. The rate veins remains controversial. About 15% of untreated isolated calf-vein
of DVT or PE within three months of superficial thrombophlebitis DVT will extend into the popliteal vein or more proximally. Anticoagu-
is about 3%, so patients with superficial vein thrombosis should be lation lowers the rate of proximal propagation and embolization.19 How-
scheduled for a repeat ultrasound in 7 days to rule out progression. ever, it is not known whether the benefit of anticoagulation outweighs
Based on the results of a large randomized controlled trial, patients the risk, nor is it known whether anticoagulation is superior to a strategy
of serial venous ultrasound with anticoagulation reserved for DVT that Disposition
propagate proximally. Several factors should lead the emergency clini- Most patients with acute DVT can be discharged from the ED as long
cian to favor anticoagulation, including: ongoing thrombotic risk (e.g., as reliable systemic anticoagulation can be established, including
active cancer, immobility), severe symptoms, DVT longer than 5 cm, ensuring that the patient can obtain and pay for their anticoagulant,
DVT close to proximal veins, or a history of prior VTE.11 High bleeding and follow-up with primary care, an anticoagulation clinic, or other
risk favors surveillance without anticoagulation. When treatment with appropriate specialty is available. Protocols that use monotherapy with
anticoagulation is used for isolated calf vein thrombosis, the dosing reg- a DOAC can facilitate outpatient management.27,28 Patients without
imen is the same as for proximal DVT (see Table 74.3). reliable follow-up or with challenges to complying with a treatment
regimen should have care coordination and education in the ED,
Phlegmasia Cerulean and Alba Dolens
observation unit, or inpatient floor prior to discharge. Patients with
Massive iliofemoral vein occlusion results in swelling of the entire leg, with severe pain may require admission. Patients with iliofemoral DVT are
extensive vascular congestion and associated venous ischemia, producing at higher risk for PE, especially when mobile thrombus is visualized
a painful cyanotic extremity. There may be associated arterial spasm result- on ultrasound, and admission is generally recommended for these
ing in phlegmasia alba dolens (painful pale leg), which may mimic an acute patients as well. Patients with possible phlegmasia should be admitted.
arterial occlusion. Elevated compartment pressures can also lead to limb
ischemia. Phlegmasia is a limb-threatening emergency. Early diagnosis
and prompt treatment with catheter-directed thrombolysis, percutaneous PULMONARY EMBOLISM
thrombectomy, or open surgical thrombectomy may be limb-salvaging.
These procedures all require interventional-suite or operating-room capa- Clinical Features
bilities. Therefore, emergency clinicians caring for patients with evidence Symptoms
of phlegmasia in hospitals without these resources immediately available The clinical presentation of PE can vary from asymptomatic to sudden
should transfer to a capable center as soon as possible. cardiovascular collapse. In fact, virtually any ED visit related to short-
ness of breath, chest pain, palpitations, dizziness, syncope, fatigue,
Upper Extremity Venous Thromboses weakness, nonspecific malaise, or functional deterioration could rep-
DVTs of the upper extremity have become more common in associa- resent a potential PE. However, this does not mean that every patient
tion with the increased use of indwelling venous catheters and wires with these symptoms should be evaluated for PE. Each patient’s symp-
for electronic cardiac devices. About half of all upper extremity DVTs toms need to be considered in the context of the entire clinical picture,
are associated with an indwelling catheter, with peripherally inserted including the likelihood of alternative diagnoses.
central catheters (PICCs) imposing the highest risk. Upper extremity The most common symptom of PE is dyspnea. If asked in a
DVT, especially axillary-subclavian vein thrombosis, occurring in the detailed way, 75% to 80% of patients with PE will report some sen-
dominant arm of relatively young, physically active, patients is called sation of dyspnea. Conversely, this means that the most common
Paget-Schroetter syndrome or “effort DVT.” This is a form of thoracic symptom of PE is absent in nearly one-fourth of patients, so clini-
outlet syndrome and may be secondary to compression of venous cians should be wary of ruling out PE simply because the patient does
structures by hypertrophied scalene muscles or a cervical rib. While the not have dyspnea (or any particular symptom). A patient with PE
data are not as robust as for lower extremity DVT, a similar diagnostic typically presents with 2 to 3 days of constant or worsening shortness
approach that combines pretest probability assessment, D-dimer test- of breath, though symptoms can be sudden. The dyspnea may be con-
ing, and ultrasound appears to exclude upper extremity DVT safely.24 stant or intermittent and perceived only with exertion. It may also be
Upper extremity DVT can cause PE, and all patients with DVT above quite vague and described as fatigue or a feeling of not being able to
the elbow require definitive treatment with anticoagulation prescribed at take a complete breath.
the same doses as lower extremity DVT (see Table 74.3). Acute PE from an Chest pain is the second most common symptom of PE. When
upper extremity DVT occurs in 5% to 8% of patients, although the PE sec- chest pain is present, it may be described as pleuritic, achy, dull, or
ondary to upper extremity DVT tend to be less severe. The recommended only as a vague discomfort. As many as one-third of patients with PE
duration of anticoagulation for upper extremity DVT remains variable, but have no chest pain. Pleuritic pain is a severe, sharp pain that repro-
most published guidelines recommend at least 3 months.11 In the absence ducibly halts respiration. When asked, many patients say that their
of pain or infection, catheter-associated DVT does not automatically war- pain is worse with deep breathing, but this is not necessarily pleuritic
rant catheter removal. The ability to anticoagulate the patient and the clinical pain. While pleuritic pain is classically described, it is only present in
necessity of the catheter should be considered. Optimal treatment of isolated 20% of patients with PE and is typically only associated with pleural
brachial vein thrombosis, often the result of a recent intravenous infusion inflammation from peripheral PE that causes pulmonary infarction
(infusion phlebitis), is uncertain. No study has demonstrated clear benefit and inflammation.
for systemic anticoagulation of brachial vein thrombosis. It is reasonable to When PE causes pulmonary infarction, symptoms can be clinically
treat these clots similarly to superficial thrombophlebitis of the leg. similar to lobar pneumonia, including pleuritic chest pain, cough,
hemoptysis, and fever. However, fever due to PE is typically low grade,
Complications and a temperature greater than 101.5°F (38.6°C) suggests infection
The most feared complication of DVT is fatal PE. However, DVT also rather than infarction. Moreover, the presence of several days of a
damages venous valves, leading to venous insufficiency, varicose veins, cough productive of sputum suggests pneumonia rather than PE.
and the post-thrombotic syndrome in 20% to 50% of patients.25 The Unexplained, unilateral leg swelling is uncommon (<30%) in
post-thrombotic syndrome includes persistent swelling, pain, burning patients with PE, but when present along with dyspnea or chest pain,
sensation, varicose veins and skin changes, and nonhealing ulcers in should raise suspicion for PE because this symptom is relatively specific.
5% to 10% of cases.25 Compression stockings may provide pain relief Overall, fewer than 5% of patients who present with syncope have
for patients with post-thrombotic syndrome, but for patients with a PE.29-31 However, unexplained syncope, especially in a patient with PE
new diagnosis of DVT, compression stockings do not reduce the inci- risk factors, should be considered as a potential presentation of PE even
dence of post-thrombotic syndrome.20,26 in the absence of dyspnea or chest pain.32
Physical Examination
Most patients with PE have no obvious abnormality on physical exam-
ination. The only finding that reliably increases the probability of a PE
diagnosis is evidence of a DVT, such as unilateral leg asymmetry, uni-
lateral edema, or tenderness along a deep vein.1 On the other hand,
the physical examination can provide clues to alternative diagnoses.
Wheezing, or a prolonged expiratory phase on lung auscultation, sug-
gests the alternative diagnosis of bronchospasm, which reduces the
probability of PE. Bilateral rales suggest the diagnosis of left ventric-
ular failure, although localized rales can also be heard over infarcted
lung tissue.
Some clinical findings can also make the diagnosis of PE more dif-
ficult. A diagnosis of PE is more likely to be delayed in patients with
baseline dementia, altered mental status, and multiple comorbidities.
Additionally, symptoms, signs, and radiographic findings that support
Fig. 74.7 Clinical autopsy photograph from showing a pulmonary embo-
lism completely occluding the right ventricular outflow system.
a diagnosis of pneumonia can lead to a missed or delayed PE diag-
nosis. Emergency clinicians may be prone to miss small, distal PE
that produce pleuritic pain, rales, and peripheral densities on chest
In its most extreme form, PE can obstruct the entire right ventric- radiographs.
ular outflow system (Fig. 74.7), leading to sudden cardiovascular col-
lapse and cardiac arrest. Approximately 25% of sudden cardiac deaths Differential Diagnosis
are thought to be due to PE. Prior to cardiac arrest, patients with PE The differential diagnosis for PE includes any thoracic process associ-
typically have a heart rate higher than their systolic blood pressure (i.e., ated with chest pain, dyspnea, or lightheadedness. Alternative diagno-
shock index >1),33 overt respiratory distress, syncope or seizure-like ses include pneumonia, acute coronary syndromes, aortic dissection,
activity, and significant anxiety. With impending cardiac arrest some pericarditis, pleural or pericardial effusion, pulmonary hypertension,
patients manifest slow agonal rhythms, possibly due to septal wall ten- pneumothorax, acute decompensated congestive heart failure, asthma,
sion leading to ischemia on the atrioventricular node and infranodal chronic obstructive pulmonary disease, gastroesophageal reflux, dys-
conducting pathways. Pulseless electrical activity ([PEA], >20 depo- pepsia, musculoskeletal pain, and nonspecific chest pain. Most alter-
larizations/min, without palpable pulses) is the most common elec- native diagnoses can be ruled out with a thorough history, physical
trocardiographic (ECG) finding in patients with cardiac arrest due examination, chest radiographs, ECG, cardiac troponin testing, and
to PE. PEA manifests from right ventricular outflow obstruction and echocardiography.34,35 In ED patients evaluated for possible PE, pneu-
impaired right ventricular contractility. The survival rate from cardiac monia is the most common alternative diagnosis suspected by clini-
arrest from PE is about 20%, even if the arrest is witnessed and bolus cians, and the most common alternative diagnosis found on computed
fibrinolytic medication is given. tomography pulmonary angiography (CTPA), present on 5% to 10%
of scans.
Vital Signs
The most common vital sign abnormality in PE is tachycardia, and Diagnostic Testing
about half of all patients with PE have a heart rate greater than 100 Figure 74.8 provides an algorithmic approach to PE diagnosis.
beats/min. Tachycardia may only be present or may be significantly
worsened with ambulation. Tachycardia is associated with more severe Bedside Assessment
PE and worse prognosis. About half of patients have an elevated respi- The first step in the evaluation of possible PE is determining whether test-
ratory rate (>20 breaths/min), though respiratory rate measurement is ing for PE is indicated at all. Each year, more than 16 million patients, or
often variable in the ED. Hypotension (systolic blood pressure < 90 mm 12% of all patients who present to the ED, have chest pain or dyspnea.
Hg) occurs in about 10% of patients but is the single most important Not all of these require an evaluation for PE. Most ED patients have at
predictor of PE mortality, associated with a four-fold increase in risk of least one risk factor for PE (e.g., advanced age, obesity, hormone use), so
death.33 PE should be on the differential diagnosis for any patient with epidemiologic risk factors for PE do not necessarily mandate a workup for
hypotension unexplained by an initial ED evaluation. Hypotension PE. Emergency clinicians currently evaluate 1% to 2% of all ED patients
from PE is due to compression of the left ventricle by the deviated right for PE with CTPA. Even more patients undergo D-dimer testing, the over-
ventricular septum. Left ventricular compression leads to decreased use of which has been shown to increase the use of CTPA. Because CTPA
filling, decreased cardiac output, poor coronary perfusion, and a spiral imposes risks to patients, including exposure to ionizing radiation, IV con-
of cardiac failure. Most studies evaluating the prognostic value of ED trast, and the risk of a false-positive interpretation, it is essential to weigh
vital signs use the patient’s worst vital signs (e.g., lowest blood pressure, the risk of a PE diagnosis against the potential risks of testing. The decision
highest heart rate). Significantly, normalization of a vital sign does not to pursue a diagnosis of PE should be based on each patient’s presentation.
reduce the probability of PE. Determining whether objective testing is indicated requires the cli-
Pulmonary embolism can produce hypoxemia (pulse oximetry nician to estimate the patient’s probability of a PE diagnosis. The “test
reading <95% at sea level or <92% at higher elevations), but the degree threshold” represents the point below which testing is more likely to
of hypoxemia is unpredictable. Approximately half of all patients with harm than benefit the patient. For PE, the test threshold is from 1.5%
PE have no evidence of hypoxemia, so a normal oxygen saturation, to 2%. Patients with less than 1.5% to 2% probability of PE are more
although reassuring, cannot rule out PE. However, hypoxemia that likely to be harmed by testing than to benefit, so testing is best avoided
cannot be explained by a known disease process increases the likeli- in these very low pretest probability patients. Conversely, patients with
hood of a PE diagnosis. When PE is diagnosed, the severity of hypox- greater than 2% probability of PE are more likely to benefit from test-
emia represents a significant independent predictor of outcomes. ing for PE than to be harmed. These patients should undergo objective
Assess pretest probability of PE
Low or intermediate High

(gestalt <40%, Wells ≤ 6, Geneva <10) (gestalt >40%, Wells >6, Geneva 10+)
D-dimer
– +
Chest imaging
CTPA Indeterminate V/Q
Normal High
– +
No PE PE
Fig. 74.8 Flowchart algorithm showing the diagnostic approach to a patient with possible pulmonary
embolism. This algorithm starts with the assessment of pretest probability and includes a combination of
D-dimer testing and pulmonary vascular imaging. In most cases, computed tomography pulmonary angiography
(CTPA) will be the imaging test of choice, but ventilation/perfusion (V̇/ Q̇) scanning is also appropriate for
patients with normal chest radiographs and a contraindication to CTPA. PE, pulmonary embolism; −, test
negative; +, test positive; CTPA, computed tomography pulmonary angiography; V̇ / Q̇ , ventilation/perfusion.
testing for PE unless there is a viable alternative diagnosis, as the pres- Biomarkers including cardiac troponin, B-type natriuretic peptide
ence of an alternative diagnosis significantly decreases the likelihood (BNP)/N-terminal pro-B-type natriuretic peptide (NT-pro-BNP), and
of PE in an ED patient. lactate can all be elevated in PE. An elevated troponin indicates myo-
It follows then that tests that demonstrate alternative diagnoses can cardial injury and is associated with higher mortality after PE.37 The
affect the need for a PE workup, and most patients who present with degree of troponin elevation predicts the degree of RV strain and mor-
symptoms suggestive of PE will have at least some testing that provides tality from PE. The appropriate use of high-sensitivity troponin assays
information on alternative diagnoses. Chest radiography, although is not well defined for PE.38 BNP and NT-pro-BNP are released from
nonspecific for PE, can provide evidence of pneumonia, congestive cardiac myocytes in the setting of myocardial distention and increased
heart failure, or pneumothorax. However, chest radiographs must be pressure. Individuals with PE and high BNP or NT-pro-BNP levels are
interpreted carefully, as pulmonary infarction may be visible on chest at increased risk of a complicated in-hospital course and 30-day mor-
x-ray as a pleural-based, wedge-shaped infiltrate, “Hampton’s hump,” tality after PE. However, isolated elevated BNP is a nonspecific finding,
which can be confused for peripheral pneumonia. and these results must be interpreted in the context of the complete
Like a chest x-ray, a 12-lead ECG provides more information about clinical picture. PE can cause end-organ hypoperfusion, and normo-
the presence of alternative diagnoses (e.g., pericarditis, cardiac isch- tensive PE patients with elevated plasma lactate (≥2 mmol/L) have
emia) than the presence of PE. When PE does cause ECG changes, it is significantly higher mortality, are more likely to develop of shock or
usually a result of acute or subacute pulmonary hypertension and asso- hypotension, require intubation, or CPR.39
ciated right ventricular dysfunction. The most common finding of PE Bedside ultrasound can be useful when PE is suspected. The pres-
on the ECG is tachycardia. Symmetric T-wave inversion in the anterior ence of acute DVT on bedside venous ultrasound in a patient with new
leads (V1–V4), the McGinn-White S1Q3T3 pattern, and incomplete symptoms of PE is equivalent to a PE diagnosis and sufficient to ini-
or complete right bundle branch block (Fig. 74.9) indicate acute cor tiate treatment. Emergency clinician–performed venous ultrasound is
pulmonale and are associated with more severe PE.36 Although these 86% to 96% sensitive and 93% to 97% specific compared to radiology-
findings are nonspecific, especially in the presence of preexisting lung performed ultrasound.40 However, emergency clinicians must keep in
disease, the presence of any one of them approximately doubles the mind that fewer than half of patients with PE have no ultrasound evi-
probability of PE in a symptomatic patient. dence of DVT, so a negative ultrasound is not sufficient to rule out PE.
Fig. 74.9 Electrocardiogram from a patient who presented with PE in the right main pulmonary artery (see
Fig. 74.3B). The presence of deep T-wave inversions in leads V1–3, a new complete right bundle branch block,
and the S1Q3T3 pattern all indicate right ventricular dysfunction (acute cor pulmonale) and are associated
with poor prognosis.
TABLE 74.4 Wells Score for Pulmonary TABLE 74.5 The Revised Geneva Score for
Embolism Pulmonary Embolism
Clinical Feature Score Clinical Feature Score
Previous PE or DVT 1.5 Age >65 years old 1
Heart rate > 100 beats per min 1.5 Previous PE or DVT 3
Recent surgery or immobilization (within 4 weeks of presentation) 1.5 Recent surgery or immobilization (within 4 weeks of presentation) 2
Clinical signs of DVT (swelling or tenderness on palpation of the 3 Active cancer (treated within the previous 12 months or currently 2
calf) receiving palliative treatment)
Hemoptysis 1 Unilateral leg pain 3
Active cancer (treated within the previous 6 months or currently 1 Hemoptysis 2
receiving palliative treatment) Heart rate <75 = 0
An alternative diagnosis is less likely than PE 3 75–94 = 3
>95 = 5
A score <2 indicates that the probability of deep vein thrombosis is low,
2–6 intermediate, and >6 high. Pain on palpation and unilateral edema of the leg 4
PE, Pulmonary embolism; DVT, deep vein thrombosis. A score <4 indicates that the probability of deep vein thrombosis is low,
4–10 intermediate, and >10 high.
Point-of-care echocardiography showing RV dilatation (a ratio of the PE, Pulmonary embolism; DVT, deep vein thrombosis.
RV to LV internal diameter in diastole [RV : LV ratio] >1 : 1) should
also increase the clinical suspicion of PE, especially in a patient with no the predictive value of the Wells score and may introduce intrarater
previous lung disease. variability.
The Revised Geneva Score (and a simplified version of the score),
Pretest Probability Assessment includes only objective elements. Patients with 0 to 3 points are defined
Once the decision to initiate a PE evaluation has been made, the next as low probability, 4 to 10 points as intermediate probability, and
step is determining the patient’s PTP of PE. The PTP guides what (if greater than 10 points as high probability. Several studies have found
any) objective testing for PE should be performed and whether empir- the Wells score and Geneva score perform similarly in clinical practice,
ical therapy should be initiated. There are several acceptable methods and both of these systems can be built into clinical decision support
for estimating PTP. An experienced emergency clinician can accurately systems to guide testing. Importantly though, when used alone, none
assess PTP using their clinical experience, or “gestalt.”35 In fact, clinical of these methods reproducibly identify the very low-risk population
gestalt has been shown to be the most accurate method of PTP assess- whose PTP lies below the 2% test threshold.
ment. Clinicians can also use one of several clinical scoring systems.
The Wells Score (Table 74.4) and the Revised Geneva Score (Table 74.5) Pulmonary Embolism Rule-Out Criteria
are both well-validated. To identify the patients in whom PE can be safely excluded without
For the purposes of determining further testing, the Wells score can objective testing, emergency clinicians should use the PE rule-out cri-
be divided into non-high (0–6 points) or high (>6 points) probability. teria, or “PERC rule” (Box 74.6).41-43 When the clinician’s unstructured
A three-tiered interpretation of the Wells score is also accepted, but PTP assessment (i.e., clinical gestalt) for PE is low, and each of the eight
the intermediate-risk middle category is less easily adapted to binary elements of the rule is satisfied, the PERC rule identifies patients whose
ED decision making. The Wells score includes a subjective question probability of PE is below the 2% test threshold. For these patients, lab-
that asks whether PE is the most likely diagnosis. This empowers the oratory testing or imaging for PE is more likely to expose the patient to
clinician to include elements that may be uncommon yet important harm than to provide benefit, so testing should not be performed. The
to consider in the PTP assessment, such as inherited clotting disor- PERC rule has been validated in large studies in the United States and
ders. However, the subjective component seems to power most of Europe.41,43 When calculating the PERC, it is important that PTP be
approach was associated with no missed PE diagnoses on follow-up

BOX 74.6 Pulmonary Embolism Rule-Out
and reduced the need for chest imaging by 17%.
Criteria (PERC Rule) The YEARS algorithm is another approach that uses three questions
• Age < 50 from the Wells score (clinical signs of DVT, hemoptysis, alternative
• Pulse < 100 diagnosis less likely than PE) to identify low PTP patients for whom
• Sao2 > 94% the higher D-dimer threshold can be used.46 YEARS has been validated
• No unilateral leg swelling in a large multicenter study and decreased the need for CTPA by 14%.47
• No hemoptysis However, in the validation study, the same result was achieved using
• No recent trauma or surgery only the subjective question of the Wells score, “Are alternative diagno-
• No prior PE/DVT ses less likely than PE?,” as the criteria for low PTP. The YEARS criteria
• No hormone use have also been adapted for use in pregnant women.48 In a study of 512
pregnant women suspected of having PE, CTPA was avoided in 195
(39%) patients, and 20 women were diagnosed with PE. However, the
efficiency of the score varied significantly according to the trimester of
determined using the clinician’s gestalt, as the PERC rule contains sim- pregnancy. Imaging was deemed not indicated for 65% of women in
ilar elements to both the Wells and Geneva scores, so may not perform the first trimester of pregnancy, but for only 32% in the third trimester.
as well in combination with these methods of PTP assessment.
Computed Tomography Pulmonary Angiography
D-Dimer Testing When the PTP is high or the screening D-dimer is positive, pulmonary
D-dimer assays are 95% to 98% sensitive and 40% to 55% specific for vascular imaging is advised. Most centers use CTPA as the primary
PE.2,44 Therefore, patients with a non-high PTP (gestalt PTP < 40%, imaging method for evaluating possible PE. CTPA is readily available,
Wells score <6, or Revised Geneva score <5) can have PE excluded with minimally invasive, and can identify alternative processes that might
a normal D-dimer concentration with negative predictive value of 99% explain the patient’s symptoms.
to 100%. Large, real-world studies have shown that many ED patients Many PE, especially central PE, can be easily visualized by the
still undergo CTPA despite a negative D-dimer. However, emergency emergency clinician. PEs appear as hypodensities in contrast-filled
clinicians should recognize that a false-positive CTPA is actually more pulmonary arteries. Most PE will be in the central part of the lumen
likely than a false-negative D-dimer, so ordering imaging “to be safe” near the bifurcation of a vessel. Filling defects that appear peripheral
exposes the patient to unnecessary risk. or concentric may be chronic or may represent alternative processes
When false-negative D-dimer results do occur, it may be because or artifacts. A PE that extends across the bifurcation of both main pul-
the PE is subacute or chronic. The half-life of circulating D-dimer is less monary arteries is called a saddle PE. Although they tend to be large
than 8 hours. However, the clinical relevance of the circulating half-life clots, many saddle PE are not occlusive, so patients may be minimally
is unknown, because large PE continue to release D-dimer as they lyse symptomatic. PE that extend into the pulmonary arteries may lodge in
over time. False-negative D-dimer measurements may also be seen with a main pulmonary artery, intralobar, lobar, segmental or subsegmental
severe lipemia and ongoing warfarin therapy. Isolated subsegmental PE artery (see Fig. 74.3). PE are typically described according to their most
seen on CTPA may also be associated with negative D-dimer results, but proximal location, so a clot that extends from the lobar pulmonary
in these cases the emergency clinician will need to determine whether artery into a segmental branch is described as lobar. While radiologists
the D-dimer is a false negative or the CTPA is a false positive. may describe a PE as “massive” based on the size of the clot, emer-
Many conditions elevate the D-dimer other than PE and DVT (see gency clinicians should remember that “massive PE” is a physiologic
Box 74.2). Because a positive result often necessitates diagnostic imag- diagnosis based on hemodynamic instability, and clot burden on CT
ing, emergency clinicians should consider the likelihood of a positive has only a loose association with mortality. Due to this confusion, the
D-dimer before ordering the test. This is especially true when CTPA term high-risk PE is a preferred term to describe the PEs resulting in
carries additional risk to the patient (e.g., pregnant women, renal insuf- hemodynamic instability.
ficiency) or is contraindicated (e.g., anaphylaxis with contrast dye). A technically adequate CTPA scan, performed on a multidetector row
The threshold for a negative D-dimer can vary by assay and local scanner, has sensitivity and specificity of 90% to 95%. Technical adequacy
laboratory. In many laboratories, the threshold below which D-dimer is requires more than 200 HU of contrast opacification in the main pulmo-
considered negative is 500 ng/mL. When the 500 ng/mL cutoff is used, nary artery and absence of motion artifact. Scans that are not technically
the threshold for a positive D-dimer can safely be adjusted according adequate are less sensitive and specific for PE. Although the sensitivity of
to the patient’s age using the following formula:2-5 CT can be improved slightly by extending the scan to include the leg veins
(CT venography), the technical reliability of this technique is limited and
Age × 10 ng / mL the additional radiation required is substantial, so most centers do not rou-

tinely perform “run-off ” venography. As with any test, the probability of
This strategy maintains a sensitivity greater than 95% and negative pre- PE after CTPA should be based on a combination of the patient’s pretest
dictive value greater than 98%, but increases the percentage of patients probability and the CTPA results. A negative CTPA in a patient with high
who can have PE excluded without pulmonary vascular imaging by 5% pretest probability should be interpreted with caution because the negative
to 6% overall and 10% to 20% in patients greater than 70 years old.5 predictive value in this population is only about 60%. Further testing such
The safety of this strategy has not been tested with D-dimer assays with as leg ultrasound may be indicated to rule out PE. Similarly, an isolated
abnormal thresholds different than 500 ng/mL. subsegmental filling defect on CTPA in a patient with low pretest prob-
As with DVT, emergency clinicians can also apply a Bayesian ability and no DVT on leg ultrasound is likely to be a false positive. In
approach by which the threshold for a positive D-dimer is increased these cases, treatment should be based on shared decision making with the
based on PTP. Using this strategy, patients with low PTP can have DVT patient. The increased detection of isolated subsegmental filling defects on
ruled out using a threshold of 1000 ng/mL (for a test that normally uses more thinly collimated CTPA images, combined with increased CTPA use,
500 ng/mL).45 A study of 2017 patients evaluated for PE found that this has led to the overdiagnosis of PE.
diagnostic testing for PE, pregnancy does not appear to increase the
risk of a VTE diagnosis.1 This is because emergency clinicians tend to
be conservative in their approach to the pregnant woman with signs or
symptoms of PE and have a much lower threshold to order diagnostic
testing. Most pregnant women whom emergency clinicians test for PE
have a low clinical probability, though there is significant variation in
the threshold to test because clinical decision rules, including the Wells
score and PERC rule, have not been validated in pregnant women.
VTE in pregnancy is relatively rare, so even the largest studies are
limited to a few dozen positive cases.48 Because D-dimer levels are
increased in pregnancy, the lower threshold to test employed by emer-
gency clinicians results in many pregnant women being referred for
imaging. However, pregnant women are justifiably cautious about fetal
radiation exposure and may not choose to undergo CTPA when it is
recommended. Emergency clinicians should have a predefined strategy
for diagnostic testing of pregnant women with possible PE. Guidelines
vary in their recommendations, and there is no single, data-driven
approach to testing.
Fig. 74.10 Diagnostic ventilation/perfusion (V̇/ Q̇) scan. These images
are high probability for acute pulmonary embolism based on the criteria Before initiating any testing for PE, it is prudent to explain the diag-
defined in the prospective investigation of pulmonary embolism diagno- nostic options to the patient, including the risks and benefits of dif-
sis (PIOPED). The first and third rows project the perfusion phases of ferent approaches, obtain her preferences, and document these stated
the examination, and the second and fourth rows show the ventilation preferences. Pregnant women should be given the opportunity to con-
phases. The black arrowheads point to wedge-shaped defects in the sider their testing preferences free from the bias associated with know-
perfusion images. Comparison with the corresponding ventilation view ing D-dimer or other test results.
immediately below shows relatively homogeneous scintillation activity When the decision is made to pursue testing, it is best to begin with
in the anatomic segments that lack perfusion. noninvasive, nonionizing testing. In a study of 395 pregnant women
(28 [7.1%] with PE), bilateral lower extremity ultrasound and D-dimer
Ventilation/Perfusion Scan testing were performed as first-line tests to either rule in or rule out
The V̇/ Q̇ scintillation scan remains a viable diagnostic option for VTE.49 A positive lower extremity ultrasound for DVT in a pregnant
patients with contraindications to iodinated intravenous contrast woman suspected of having PE is sufficient to confirm the diagnosis
and vulnerable kidney function. The accuracy and precision of the of PE and to initiate treatment. When the conventional threshold for a
V̇/ Q̇ scan were shown in the Prospective Investigation of Pulmonary positive D-dimer used for nonpregnant patients is applied (e.g., <500
Embolism Diagnosis (PIOPED) study, which compared the results of ng/mL), a normal D-dimer excludes PE in pregnant women as reli-
V̇/ Q̇ scanning with the most accurate criterion standard test available ably as in the nonpregnant population. However, most women will
at the time, catheter pulmonary angiography. Figure 74.10 shows the have a positive D-dimer and negative venous ultrasound results. In the
results of a high-probability V̇/ Q̇ scan. A high-probability V̇/ Q̇ scan previously mentioned study, 83% of women required CTPA or V̇/ Q̇
confirms the diagnosis of PE, and a normal scan (i.e., no perfusion scanning. Therefore, before pursuing this approach, pregnant women
defect) excludes PE. However, only one-third of V̇/ Q̇ scans fall into should be informed that imaging is likely to be needed. This will inform
either the high-probability or normal categories. The remaining two- their shared decision to pursue a PE workup.
thirds of scans are read as either low probability or intermediate proba- One reason a small proportion of women can be ruled out with
bility. These categories indicate indeterminate scans that are essentially noninvasive testing is that D- dimer levels fluctuate and steadily
nondiagnostic. Many patients, therefore, require additional diagnostic increase throughout pregnancy until they peak on the first day after
testing after an indeterminate V̇/ Q̇ scan. When feasible, these patients delivery.50 The proportion of women with normal D-dimer values
should undergo CTPA, though negative bilateral venous ultrasonog- is 50% to 85% in the first trimester, decreases to 23% to 33% in the
raphy of the legs can also be used to rule out PE in patients with a low second trimester, and to 0% to 4 % in the third trimester. Pregnancy-
PTP and a low-probability V̇/ Q̇ scan. specific threshold values for each trimester have been suggested to help
V̇/ Q̇ scanning is generally associated with lower fetal radiation increase the proportion of women who can have PE ruled out without
exposure than CTPA. When the perfusion portion of the V̇/ Q̇ scan imaging. One model suggests using a threshold of 750 ng/mL in the
is performed first, and the scan is normal, the ventilation portion of first trimester, 100 ng/mL in the second trimester and 1250 ng/mL in
the scan may be avoided, further reducing the radiation exposure to the third trimester. However, outcome data supporting this approach
the woman and the fetus. However, for technical reasons, most insti- are limited to studies with fewer than 20 pregnant women with VTE,
tutional protocols perform the ventilation portion of the scan first. In so the safety of D-dimer adjustment based on pregnancy trimester has
addition, because many women will require CTPA after an indetermi- not been validated.
nate V̇/ Q̇ scan, and will be exposed to two doses of radiation, many The YEARS approach has been modified for use in pregnancy,
institutions now recommend shielded, low- radiation CTPA rather with a screening venous ultrasound performed first, followed by PTP-
than V̇/ Q̇ scanning as the first test for pregnant women. adjusted D-dimer based on the same three Wells score criteria used
in the original study.48 Among 498 women, 4 had DVT on screening
Pregnant Women ultrasound, and 16 were diagnosed with PE based on a combination of
Pregnancy is a hypercoagulable state, and when combined with PTP adjusted D-dimer testing and chest imaging. Nearly one-third of
mechanical compression of the pelvic veins by the gravid uterus, women in the third trimester had PE ruled out without imaging, and
increases the risk of PE, especially in the third trimester and post- only one subject was diagnosed with DVT after having PE ruled out in
partum period. However, when studied in ED patients undergoing the ED. However, as with other studies of pregnant women, the small
TABLE 74.6 Fetal and Maternal Radiation Exposure Associated With Diagnostic Tests for PE
Fetal Exposure Maternal Exposure
Imaging Method (mSv) (mSv)
Chest x-ray (two view) <0.01 <0.01
Ventilation lung scan 0.1–0.3 <0.01
Perfusion lung scan 0.1–0.6 0.2–1.2
CTPA 0.01–0.66 7–70
Catheter pulmonary angiogram 3–6 15–20
CT venogram of the iliac veins 10–50 10–50
CTPA, Computed tomography pulmonary angiography; CT, computed tomography.
Adapted from: Linnemann B, Bauersachs R, Rott H, et al. Diagnosis of pregnancy-associated venous thromboembolism—position paper of the
Working Group in Women’s Health of the Society of Thrombosis and Haemostasis (GTH). Vasa. 2016;45(2):87-101.
number of PE events makes it difficult to recommend this approach pulmonary vasospasm, sudden dysrhythmias, or respiratory failure.
confidently. Of ED patients without hypotension, 1% to 3% experience cardiac
For pregnant women for whom PE cannot be ruled in with ultra- arrest while in the ED or die within 24 hours. Factors independently
sound or ruled out with D-dimer, chest imaging is indicated. In this associated with clinical deterioration after PE diagnosis include ED
case, women often have questions about radiation exposure and the hypotension, hypoxemia, prior coronary artery disease, residual deep
safety of the fetus. For comparison, the fetal radiation exposures of var- vein thrombosis, and right heart strain on echocardiogram. Clues to
ious tests for PE are presented in Table 74.6. The average dose of natu- oncoming cardiopulmonary decompensation include worsening respi-
rally occurring background radiation is 0.5 to 1.0 mSv over a normal ratory distress and hypoxemia, increasing tachycardia and shock index,
gestation. Exposure to radiation doses lower than 50 mSv has not been or a change in mental status. Deterioration in the ECG from a narrow-
shown to be associated with different pregnancy outcomes.51 In the complex tachycardia to an incomplete right bundle branch block to a
first two weeks after conception, a radiation dose of 50 to 100 mSv can complete right bundle branch block also suggests life-threatening pul-
cause the failure of blastocyst implantation and spontaneous miscar- monary hypertension.
riage. Assuming the blastocyst implants, radiation effects are unlikely
because blastocyst cells are omnipotent and damaged cells can be Management
replaced (the “all-or-none period”). The embryo is most vulnerable to The treatment of acute PE depends on the risk that the PE poses to
radiation between the eighth and fifteenth weeks of gestation. A defini- the patient (Table 74.7). PE are classified as high risk, or “massive,”
tive threshold has not been determined, but it is estimated that the risk intermediate risk or “submassive,” and low risk. These risk categories
of congenital malformations, intrauterine growth restriction, intellec- are defined by the hemodynamic status of the patient and the presence
tual disability, and pregnancy loss increase at radiation doses beyond of right ventricular dysfunction, not by the size of the thromboembo-
100 to 200 mSv. The risk of cancer associated with fetal radiation expo- lism. Therefore, to avoid confusion sometimes associated with the term
sure is difficult to predict, but even with higher radiation doses many “massive” and “submassive” we use the terms high, intermediate, and
times higher than those used for CTPA, the increase in lifetime cancer low risk.
due to in-utero radiation exposure from diagnostic imaging seems to High-risk PE result in hemodynamic instability, defined as a sys-
be negligible.51 tolic arterial pressure (SBP) less than 90 mm Hg that is sustained for
Data comparing fetal radiation exposure do not clearly favor 15 minutes and not caused by dysrhythmia or other etiology.57 Patients
V̇/ Q̇ or CTPA. Regardless of the approach, efforts should be made to who have a drop in their baseline SBP of greater than 40 mm Hg, who
minimize radiation exposure to the developing fetus. Because 99Tc is require vasopressors, or who have profound bradycardia (<40 bpm)
excreted in the urine, prehydration with intravenous saline, frequent can also be considered high-risk PE. It is also reasonable to require
postscan urination, and insertion of a urinary catheter are unproven that the patient’s PE be large enough to conceivably cause the patient’s
steps to reduce fetal exposure to radiation associated with V̇/ Q̇ scan- hemodynamic instability.
ning. When CTPA is performed, shielding of the abdomen and tube Intermediate-risk PE requires that the patient be hemodynami-
voltage modulating technology may reduce fetal radiation exposure. cally stable and have evidence of RV dysfunction on echocardiogram
Thus, while high-probability (i.e., diagnostic) or normal (i.e., exclu- or a positive troponin. Echocardiographic evidence of RV dysfunction
sionary) V̇/ Q̇ scanning may be associated with lower fetal radiation includes RV dilatation, hypokinesis, or bowing of the intraventricular
exposure than CTPA, the difference in radiation exposure of the two septum toward the left ventricle (LV). RV dilatation on CT (defined
tests is minimal. However, many women have nondiagnostic V̇/ Q̇ as an RV : LV ratio >1) is 88% sensitive for RV dysfunction on echo-
scans and need to undergo CTPA that will expose them to more radia- cardiogram, but is only 39% specific.58 CT evidence of RV dysfunc-
tion than if a CTPA had been performed first. Therefore, CTPA seems tion should, therefore, be confirmed with echocardiography. Some
the best first test for most pregnant women. guidelines also use a clinical risk model (i.e., the pulmonary embolism
severity index [PESI]) and troponin values to subcategorize patients as
Clinical Course intermediate-low or intermediate-high risk.59
The clinical course of patients with obstructive PE can be unpredict- Low-risk PE are hemodynamically stable with no evidence of RV
able. The mortality of high-risk PE is greater than 25%,52 but many dysfunction. These risk categories are important because the mortality
patients with high-risk PE remain stable in the ED after initial resus- increases from less than 1% to 3% (low risk) to 3% to 15% (interme-
citation. In otherwise stable patients, clinical deterioration can occur diate risk) to 15% to 50% (high risk). Accordingly, advanced therapies
rapidly due to embolization of clot material, release of mediators of like thrombolysis or thromboembolectomy are not recommended
TABLE 74.7 Risk Stratification and Treatment Recommendations for Acute PE

Category Criteria Action
Low-Risk PE Hemodynamically stable Initiate anticoagulation (DOAC preferred)
No right ventricular dysfunction Safe for outpatient treatment based on Figure 74.12 or negative Hestia
criteria
Troponin negative Optional admission to inpatient service or observation unit
Intermediate-Risk PE Hemodynamically stable Initiate anticoagulation (DOAC or heparin)
Any of the following: Thrombolysis or thromboembolectomy in a minority of cases
• R ight ventricular dysfunction on echocardiogram or CTPA Activate PERT
• T roponin positive Admission to monitored bed or ICU
• B NP/NT-pro-BNP positive
High-Risk PE Hemodynamically unstable Initiate anticoagulation (heparin if thrombolysis or thromboembolectomy
planned)
Activate PERT
Thrombolysis or thromboembolectomy unless contraindicated
Admission to ICU
PE, Pulmonary embolism; DOAC, direct-acting oral anticoagulant; CTPA, computed tomography pulmonary angiography; PERT, pulmonary embolism
response team; BNP, brain natriuretic peptide; ICU, intensive care unit.
for low-risk patients, sometimes recommended for intermediate-risk Extracorporeal membrane oxygenation (ECMO) can unload the
patients, and recommended for high-risk patients. right ventricle, increase cardiac output, and provide a bridge to throm-
Figure 74.11 presents a comprehensive management plan for diag- bolysis or thromboembolectomy for patients with high- risk PE.54
nosed PE. This algorithm includes the spectrum of treatment options, ECMO requires institutional infrastructure and expertise which is
including some that may only be available in a large tertiary care only available in specialized centers. Survival of patients with PE who
service hospital. Emergency clinicians should be familiar with the require ECMO is about 70%.54,55
treatment options where they practice. Pathways similar to this have
been adopted by multidisciplinary PE response teams (PERT).52,60-62 Pulmonary Vasodilators
At the left-most side of the algorithm, patients can be discharged to Inhaled nitric oxide and epoprostenol have been shown to decrease
home from the ED. At the right side, patients with a massive PE and pulmonary vascular resistance and pulmonary arterial pressure in case
no contraindications receive thrombolysis or are referred for emergent series. However, a clinical trial failed to show benefit of inhaled nitric
thromboembolectomy. oxide in hemodynamically stable patients with PE and right ventricular
dysfunction.56
Airway, Oxygenation, and Ventilation
Hypoxemia is common in PE and hypoxemic vasoconstriction may Standard Anticoagulation
worsen acute pulmonary hypertension, so supplemental oxygen should Patients with imaging confirming DVT or PE should receive anticoag-
target pulse oximetry above 90%. However, intubation and positive- ulation using one of the agents in Table 74.3, administered in the ED as
pressure ventilation increase intrathoracic pressure, lower preload, and soon as the diagnosis is confirmed. Patients with a high PTP, no contra-
can worsen RV compression of the LV. This may precipitate hemody- indication to anticoagulation, and evidence of hemodynamic instabil-
namic collapse and cardiac arrest in patients with severe PE. Intuba- ity, including recent syncope, hypotension, hypoxemia, or right heart
tion should, therefore, be avoided whenever possible. When intubation dysfunction, should receive empirical anticoagulation while awaiting
is necessary, the emergency clinician should optimize pre-intubation pulmonary vascular imaging. Direct-acting oral anticoagulant medica-
hemodynamics, including with vasopressors, prior to induction.53 tions apixaban or rivaroxaban are now considered first-line therapy for
most PE.11 These drugs specifically inhibit one enzyme in the clotting
Hemodynamic Resuscitation pathway, are orally available, and provide therapeutic anticoagulation
For patients who present with actual or pending hemodynamic insta- effect as rapidly as subcutaneous LMWH. Clinical trials support the
bility, the first priority is resuscitation. For hypotensive patients, small use of apixaban and rivaroxaban without prior or concomitant use of
volume boluses (such as 250–500 mL) can improve cardiac output.12 heparin. By obviating the need for twice-daily subcutaneous injec-
However, excessive administration of intravenous fluids can worsen tions and blood monitoring, these drugs can also facilitate outpatient
RV distension and compression of the left ventricle by the intraventric- treatment of PE and DVT.28 DOAC medications have not been exten-
ular septum. Figure 74.5 shows that with increased pulmonary artery sively tested in patients with severe PE, pregnant women with PE, or
pressures, the dilated RV distends towards the intraventricular septum, in patients with hypercoagulable states like antiphospholipid anti-
which can decrease left ventricular volume. This, in turn, decreases car- body syndrome. For these patients, LMWH is preferred. Compared to
diac output, increases myocardial ischemia, and can precipitate cardiac unfractionated heparin, meta-analyses show that LMWH is associated
arrest. with lower rates of major hemorrhage, heparin-induced thrombocy-
If the administration of intravenous fluids is contraindicated or topenia, and recurrent VTE, and has similar cost. Most hematologists,
no longer beneficial, vasopressors should be administered, and nor- internists, and obstetricians prefer that pregnant patients with VTE
epinephrine should be the initial agent in blood pressure support.12 receive twice-daily LMWH. Both DOAC and LMWH effects can be
Dobutamine is useful as an adjunct but may worsen hypotension unreliable in patients with severe renal impairment. For these patients,
unless coadministered with norepinephrine. unfractionated heparin is usually preferred. Patients with a history of
PE diagnosed
Positive CTPA, high-probability V/Q
Risk stratify
Vital signs, echocardiogram, CTPA, biomarkers
Intermediate risk
Low risk High risk
Normotensive,
Normotensive, no Proximal clot,
proximal clot,
RV dysfunction hypotension
RV dysfunction
Cardiac arrest?
No Yes
Anticoagulation Anticoagulation Anticoagulation

DOAC DOAC, or heparin IV heparin
Activate Activate
PERT PERT
Intermediate- Intermediate- Contraindications to

low risk high risk thrombolysis?
Contraindications
None Absolute
to thrombolysis?
Relative
No Yes
IVC filter CDT or Catheter

if residual Systemic low-dose or
CDT
proximal thrombolysis systemic surgical
DVT thrombolysis thrombectomy
Persistent hemodynamic
instability?
No Yes
ECMO
Outpatient treatment Admit to inpatient floor
Admit to ICU
with follow-up <7days or observation unit
Fig. 74.11 Flowchart algorithm showing the treatment approach to patients diagnosed with acute pulmonary embolism (PE) in the emergency depart-
ment of a full-service, tertiary-care hospital. This algorithm focuses on PE-specific treatment, rather than acute resuscitation. Low-risk PE are hemo-
dynamically stable with no evidence of RV dysfunction. Intermediate-risk PE are hemodynamically stable but have evidence of RV dysfunction on
echocardiogram or a positive troponin. Echocardiographic evidence of right ventricular (RV) dysfunction includes echocardiographic RV dilatation, hypo-
kinesis or bowing of the intraventricular septum towards the left ventricle (LV) or RV : LV ratio >1 on CT. High-risk PE are hemodynamically unstable,
defined as a systolic arterial pressure (SBP) < 90 mm Hg that is sustained for 15 minutes and not caused by a dysrhythmia or other etiology, a drop in
baseline SBP of > 40 mm Hg, vasopressors, or profound bradycardia (<40 bpm). For dosing of anticoagulation, see Table 74.3. Pulmonary embolism
response teams (PERT) can expedite the care and access to advanced therapies for patients with intermediate-and high-risk PE. Absolute contraindi-
cations to thrombolysis include: (1) gastrointestinal bleeding within previous 30 days; (2) active hemorrhage in any of the following sites at the time of
enrollment—intraperitoneal, retroperitoneal, pulmonary, uterine, bladder, or nose; (3) head trauma causing loss of consciousness within previous 7 days;
(4) any history of hemorrhagic stroke; (5) ischemic stroke within the past year; (6) history of intraocular hemorrhage; (7) known or suspected intracranial
metastasis; (8) liver failure with prothrombin time abnormal (international normalized ratio [INR] > 1.7); (9) surgery that required opening of the chest
cavity, peritoneum, skull, or spinal canal within the previous 14 days; (10) subacute bacterial endocarditis under treatment; (11) pregnancy; (12) large
pericardial effusion. Relative contraindications to thrombolysis include: age > 75 years; dementia; surgery more than 30 days but less than 60 days prior;
any prior stroke; symptoms suggesting transient ischemic attack in the past 30 days; any prior gastrointestinal bleeding; concurrent use of a thienopy-
ridine (e.g., clopidogrel); INR > 1.7 from warfarin use; any metastatic cancer, recent fracture, recent fall with head strike, history of hematuria, recent
dental extraction, or orthopedic surgery. Systemic thrombolysis dosing regimens: alteplase (recombinant tissue plasminogen activator, rtPA) 100 mg IV
over two hours, reteplase 10 units IV over two minutes and then repeated 30 minutes later or tenecteplase as a single weight-based bolus dose over 5
to 10 seconds. Low-dose thrombolysis dosing regimen: alteplase 25–50 mg IV over two hours. PE, Pulmonary embolism; CTPA, computed tomography
pulmonary angiography; V̇ / Q̇ , ventilation perfusion; RV, right ventricle; DOAC, direct acting oral anticoagulant; IV, intravenous; CDT, catheter-directed
thrombolysis; DVT, deep vein thrombosis; ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit.
patients remain subtherapeutic throughout the first 24 to 48 hours of

PE treatment.63 As soon as it is decided that the patient will not receive
thrombolysis or thrombectomy, it is best to convert to LMWH or a
Yes
DOAC.
Social or psychological barriers to
outpatient treatment? Anticoagulation therapy for patients with isolated subsegmental PE
No is controversial. If a patient has no evidence of DVT on bilateral lower
extremity ultrasonography, no signs of right heart dysfunction, and no
Yes
Abnormal vital signs? ongoing major risk for thrombosis (e.g., active malignancy, cast immo-
bilization), it is reasonable to withhold anticoagulation. The decision
No
to proceed with anticoagulation in this situation should be guided by
Coronary artery disease Yes individual patient risk profiles and preferences.10,11
congestive heart failure?
Reversal of Anticoagulation. All anticoagulants commonly used
No in the ED, including the DOACs, now have effective reversal agents.
Elevated troponin? Yes Admit The anticoagulant effect of unfractionated heparin can be almost
Creatinine clearance <30 ml/min. completely and rapidly reversed with protamine sulfate, whereas
No LMWH can only be 50% neutralized with protamine. Protamine has
High bleeding risk Yes no effect on fondaparinux, rivaroxaban, or apixaban. Andexanet alfa,
with anticoagulation? a modified recombinant inactive form of human factor Xa, is now
No No No available to reverse anticoagulant effect (anti-factor Xa activity) in
patients with life-threatening or uncontrolled bleeding on rivaroxaban
Small Intermediate Large
and apixaban.64,65 Although most patients treated in the single-arm
PE PE PE
trial achieved good hemostasis, the lack of a comparison (e.g., placebo)
No
group makes it difficult to assess the effectiveness of andexanet alfa in
Echocardiogram or CTPA Yes reducing clinically important bleeding. The effectiveness of factor Xa
with right heart strain? inhibitor reversal using four-factor activated prothrombin complex,
No 50 U/kg IV, is limited and one meta-analysis concluded that these
Venous ultrasound with Yes treatments may not result in more rapid reversal of anticoagulation
iliofemoral DVT? than simply discontinuing the DOAC.66-71
No
Thrombolytic (Fibrinolytic) Therapy
Discharge Thrombolytic therapy in PE remains a controversial treatment. It is
Fig. 74.12 Flowchart showing a validated algorithm for the identifica-
generally agreed that patients with arterial hypotension (systolic blood
tion of low-risk patients with acute PE who are candidates for outpatient pressure < 90 mm Hg or > 40-mm Hg drop from baseline) should
treatment and discharge from the emergency department (or ED obser- receive thrombolysis. Appropriate regimens for full-dose thromboly-
vation unit). Key to this approach is the availability of early follow-up sis include alteplase (recombinant tissue plasminogen activator, rtPA)
and the ability of patients to obtain their prescribed anticoagulants after given as a 100-mg IV bolus over two hours, reteplase given as 10 units
discharge. Abnormal vital signs include any hypotension or hypoxemia IV over two minutes and then repeated 30 minutes later, and tenect-
requiring oxygen therapy. Patients may be discharged with mild tachy- eplase given as a single weight-based bolus dose over 5 to 10 seconds.
cardia, but patients with severe tachycardia should be observed in the Practically speaking, two IV boluses of 50 mg separated by 15 min-
hospital. Renal insufficiency (creatinine clearance < 30 mL/min) is not utes may be more realistic than a two-hour infusion of alteplase for an
itself an indicator of higher risk PE, but severe renal impairment is a
unstable patient. We recommend thrombolytic therapy for high-risk
contraindication for both direct-acting anticoagulant (DOAC) and low-
molecular- weight heparin (LMWH) anticoagulation. PE, pulmonary
PE patients without contraindications, but specific regimens will vary
embolism; CTPA, computed tomography pulmonary angiography; DVT, by institution, availability, and pharmacy protocols.
deep vein thrombosis. For patients with intermediate-risk PE, data and clinical guidelines
are conflicting. The PEITHO study randomized 1005 patients with
heparin-induced thrombocytopenia should receive fondaparinux, arg- intermediate-risk PE, defined as right ventricular dysfunction on echo-
atroban, apixaban, or rivaroxaban. cardiogram and a positive troponin. Thrombolysis with tenecteplase
For patients in whom thrombolysis, surgery, or another advanced was associated with significant reduction in the composite endpoint of
intervention is being considered, intravenous unfractionated heparin clinical deterioration and 7-day mortality, but no decrease in mortality
may be preferable because it has a short half-life, and anticoagulation alone. There was, however, a significant increase in risk of intracranial
should be discontinued prior to thrombolysis or a procedure. However, hemorrhage, particularly in patients more than 65 years old. Subse-
intravenous heparin provides unreliable anticoagulation, and many quent meta-analyses of randomized trials that compared thrombolysis
plus heparin to heparin alone have reached different conclusions about ovale.51 The perioperative mortality is highest for patients who require
mortality benefit, with one study finding significant improvement72 CPR before surgery, and lowest for stable patients who have an inferior
and another no difference.73 This difference seems to be mostly due to vena cava (IVC) filter placed before surgery, so emergency physicians
the definition of intermediate-risk PE used to characterize studies in should involve an experienced cardiothoracic surgeon as early as pos-
the meta-analyses. sible in the care of a potential thromboembolectomy patient. Surgical
The benefit of thrombolysis for long-term outcomes in patients with embolectomy may be the best option for patients who have severe PE
intermediate-risk PE is also questionable. In the TOPCOAT trial of 88 with a contraindication to fibrinolysis; however, extracorporeal per-
patients with intermediate-risk PE, thrombolysis improved functional fusion requires intensive heparin anticoagulation, and the patient’s
capacity and subjective perception of wellness. However, in PEITHO, mental status cannot be monitored during surgery—a key concern
thrombolysis had no effect on long-term mortality, residual dyspnea, in patients with a high risk of intracranial hemorrhage. The admin-
or RV dysfunction.74 Overall, we do not recommend full-dose intrave- istration of fibrinolytic therapy does not absolutely preclude surgical
nous thrombolysis for intermediate-risk PE except in carefully selected intervention. Patients treated with a fibrinolytic agent can undergo
patients at high risk for clinical deterioration and low risk of bleeding. sternotomy or thoracotomy for thromboembolectomy and survive
Intravenous thrombolysis for low-risk PE or DVT without PE is not without fatal hemorrhage, but the decision to perform an open throm-
indicated. boembolectomy ultimately resides with the cardiothoracic surgeon.
Thrombolytic regimens that use a lower dose (i.e., 50 mg of alteplase)
thrombolysis have also been examined to determine whether lower Inferior Vena Cava Filters
thrombolytic doses reduce major bleeding associated with thrombol- For the vast majority of ED patients, placement of an inferior vena cava
ysis. While studies report low rates of intracranial hemorrhage, meth- (IVC) filter is not indicated. However, for a patient diagnosed with
odological issues and the outcome definitions used in the studies have PE in the presence of an absolute contraindication to anticoagulation,
tempered enthusiasm for this approach until further data are available. such as a recent cerebral hemorrhage, large cerebral infarction, or brain
Empirical thrombolysis has not shown benefit in undifferentiated metastases, the appropriate consultant should be contacted for urgent
cardiac arrest, but thrombolysis is recommended in cardiac arrest due placement of an inferior vena cava filter. Patients with a central PE who
to presumed PE, when there are no apparent contraindications. Both are unlikely to survive embolization of a proximal (e.g., iliofemoral),
alteplase and tenecteplase may be given as a 50-mg bolus. To promote mobile DVT should also be referred for urgent IVC filter placement.
circulation of the thrombolytic medication, CPR should be continued The judgment that a patient is unlikely to survive an additional embo-
for at least 60 minutes before terminating resuscitation.75 lization is based on clinician experience and should be made alongside
The primary risk of thrombolysis is bleeding. Studies suggest rates of interventional radiology or cardiology consultants.
major bleeding between 9% to 22%, and intracranial bleeding between
1.5% to 3%. The risk of intracranial hemorrhage appears to be highest Pulmonary Embolism Response Teams (PERT)
in patients older than 65 years. One clinical prediction score for intra- PERTs are rapid response teams composed of multidisciplinary spe-
cranial hemorrhage after thrombolysis for PE includes peripheral vas- cialists, including emergency physicians, that can be mobilized to
cular disease, age greater than 65 years, prior cerebrovascular accident, provide real-time expertise and expedite the care of patients with life-
and prior heart attack, though this study awaits external validation.76 threatening PE. Approximately 60% of PERT activations originate in
the ED.58 PERTs have been shown to facilitate access to advanced ther-
Catheter-Directed Interventions apies such as catheter-directed interventions, surgery, and extracorpo-
Catheter- directed interventions include catheter- directed throm- real membrane oxygenation, reduce disposition times, and improve
bolysis and percutaneous thromboembolectomy. The advantages of overall care.60,61 For patients with life-threatening PE, guidelines now
catheter-directed thrombolysis include relatively low doses (4–24 mg recommend the involvement of a PERT.12
of tPA over 12–24 hours) and delivery of thrombolytic medication near
or directly into the PE.77-79 Catheter-directed thrombolysis appears to Mortality and Morbidity
be as effective as systemic thrombolysis in high-risk PE with a lower For hemodynamically stable patients with PE, a multicenter registry
risk of bleeding77-79 and may be superior to anticoagulation alone in of 1880 patients diagnosed with PE in the ED found 1.1% mortality
intermediate-high risk PE, though, to date, only one small randomized directly attributable to PE and 5.4% all-cause mortality. In contrast,
clinical trial has compared catheter-directed thrombolysis to heparin the mortality of hemodynamically unstable patients is 25% to 50%.52
alone. This study of 58 patients showed catheter-directed thrombolysis Approximately half of PE survivors will have exercise intolerance and
improved RV function at 24 hours, though there was no difference at dyspnea that degrades their quality of life,80,72 and 40% to 50% of
90 days. Single-arm studies have found similar results.77-79 patients with DVT will have long-term symptoms consistent with post-
Other catheter- directed interventions include clot maceration thrombotic syndrome.25,73 The etiology of exercise limitation after PE
and suction thrombectomy. Clot maceration physically breaks up the is not well understood but may have more to do with deconditioning
clot, and may be used with thrombolysis, but can increase pulmonary than decreases in cardiac or pulmonary function.80
hypertension through the release of vasoactive mediators. Suction
thrombectomy devices mechanically withdraw clot from the pulmo-
DISPOSITION
nary artery and have the benefit of not requiring thrombolysis or anti-
coagulation. While results are promising, there is no current evidence Approximately 20% to 30% of patients with PE can safely be discharged
that catheter-directed interventions offer a survival advantage over sys- from the ED, usually with DOAC therapy and close outpatient follow
temic thrombolysis or anticoagulation alone. -up.27,28,81-84 Figure 74.12 shows a validated system for identification
and risk stratification of patients with PE. This system has been shown
Surgical Embolectomy to facilitate the outpatient treatment of up to 29% of ED patients with
Surgical thromboembolectomy can be life-saving for patients with PE, with less than 1% all-cause 30-day mortality and 0.5% 30-day major
severe refractory hypotension or free-floating thrombi in the right bleeding.28 The Hestia Criteria (Box 74.7) can also be used to iden-
heart (“clot-in-transit”), especially if the clot crosses a patent foramen tify patients who are safe for outpatient treatment.82 The pulmonary

Pulmonary Embolism and Deep Vein Thrombosis: Christopher Kabrhel

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Pulmonary Embolism and Deep Vein Thrombosis: Christopher Kabrhel

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pulmonary Embolism and Deep Vein Thrombosis: Christopher Kabrhel

Uploaded by

Copyright:

Available Formats

74